Applied Mechanics and Materials Vols. 513-517

Abstract: Cloud task scheduling is a hot technology today, how to effectively improve the utilization of resources, time efficiency, load balancing, is the focus and difficult of the study. The time efficiency, load balancing of K-Min algorithm still need to be improved, so this paper proposes cloud computing task scheduling algorithm based on modified K-Means (Improved K-Min), firstly, This paper improves the k-means algorithm using the BFA and PSO,then according to the length attribute of the task, resource requirements, the algorithm uses the improved K-means for packet processing tasks, then performs Min-Min scheduling algorithm within the group. Through theoretical research and simulation of Cloud-sim platform, when the number of tasks is 300, experimental result is best, comparing with Min-Min algorithm, the total task completion time improved 17.13%.

Abstract: As the ubiquitous penetration of high-speed Internet and mobile devices, the media consumption landscape has been significantly transformed due to the increasing popularity on social interactions and multi-screen viewing behaviors. This paper proposes a second screen solution based on Cloud Centric Media Platform. We will introduce a complete and robust layering design of Cloud Centric Media Platform, and an innovative application called video teleportation. We build a proof-of-concept testbed based on a private cloud at Nanyang Technological University (NTU). The performance evaluations indicate the application is efficient, feasible and easy to use.

Abstract: The classification performance of the classifier is weakened because the noise samples are introduced for the use of unlabeled samples in Tri-training. In this paper a new Tri-training style algorithm named AR-Tri-training (Tri-training with assistant and rich strategy) is proposed. Firstly, the assistant learning strategy is posed. Then the supporting learner is designed by combining the assistant learning strategy with rich information strategy. The number of mislabeled samples produced in the iterations of three classifiers mutually labeling are reduced by use of the supporting learner, moreover the unlabeled samples and the misclassified samples of validation set can be fully used. The proposed algorithm is applied to voice recognition. The experimental results show that AR-Tri-training algorithm can compensate for the shortcomings of Tri-training algorithm, further improve the testing rate.

Abstract: Routing strategy with effective and saving energy is an important problem in the research on the application of wireless sensor network into the farmland micro climate and soil moisture monitoring,it analyzed the disadvantage of the layered tree routing algorithm and ZigBee routing algorithms in energy saving, combined with ZigBee network topology,establishing an optimal node analysis model based on fuzzy decision, bringing forward a new routing algorithm which is suitable for large-scale farmland gradient environment, this new algorithm find out the quantitative relation between energy consumption and routing node selecting strategy,which can greatly decrease the route hop number by routing discovery mechanism and create an optimum goal group. Simulation results showed that compared with the layered tree and ZigBee routing algorithms,the new routing algorithms can significantly reduce the energy consumption of routing process.

Abstract: Based upon the intention summary of software quality management, this paper gives plain description of PDCA circulation, software quality scheme, quality control, quality assurance, and quality improvement. Meanwhile, one quality evaluation is put forward. This will benefit the software quality management of information system.

Abstract: Based upon the synopsis of software requirement development, this paper describes the practices of user requirements exploitation, product requirements exploitation, requirements analysis and validation, and then puts forward engineering process relation and requirement development practice. Meanwhile, further works are brought forward. This will benefit the software requirement development of information system.

Abstract: Labelled models are used in researching areas of communication networks, cryptography, computer science, biology, information networks etc, and they are for simulating real networks or studying algorithms.The aim of this work is to generalize several labellings to models related with complex networks.

Abstract: IIC BUS (Inter Integrated Circuit BUS) is a bus that system chip control peripherals equipment commonly used as a simple, two-way, two-wire ,and synchronous serial bus to solve the communication between these devices. The paper discuss the ways how to use the IIC serial data communication protocol according to the IIC controller features ,adopting master-slave structure to control IIC device of AT24C08 chip with S3C2410A as the carrier to fulfill data communication.

Abstract: The paper introduces the design scheme of the balanced preemptive selection system, and how to develop components implement the business logic, and deployment components for COM+. Developing high efficienctly, safe, easy maintenance of the selection course system With the Web server platform based on IIS.

Abstract: Algorithm of vehicle scheduling optimization could be integrated in the GIS platform. Therefore, distribution software can automatically make the delivery plan and managers also can make the optimizing choice of the optimal distribution route. Firstly, this paper introduces the necessity of introducing GIS into the logistics industry. Moreover, advantages and disadvantages of the current path planning in the logistics distribution which often used in some algorithms (genetic algorithm, mountain climbing algorithm, ant colony algorithm, etc.) will be listed. Finally, a more practical hybrid algorithm will be used to the GIS so that managers can optimize the logistics distribution path selection.